Confronting ISO: On the Use of the dBA Symbol

for A-Weighted Sound Levels

Enfrentando a la ISO: Sobre el uso del símbolo dBA

para los niveles sonoros con ponderación A

Federico Miyara

Laboratorio de Acústica y Electroacústica,

Universidad Nacional de Rosario

Riobamba 245 bis, 2000 Rosario, Argentina

fmiyara@fceia.unr.edu.ar

Abstract— The International Organization for Standardiza-

tion (ISO) recommends the symbol dB for all instances where

decibel units are used, particularly in the cases of quantities

such as the A-weighted and C-weighted sound pressure levels,

ruling out the symbols dBA y dBC. In this paper several

arguments are put forward as to why such recommendation is

mistaken and should be revised, taking into account that the

dB prefix is used in the symbols of many substantially different

units as well as the fact that there are examples of situations

where the criterion hereby proposed is accepted.

Keywords: ISO; units; decibel A.

Resumen— La Organización Internacional de Normaliza-

ción (ISO) recomienda el símbolo dB para todas las instancias

de uso de unidades de decibel, en particular en los casos de

cantidades como el nivel de presión sonora con ponderación A

y C, censurando los símbolos dBA y dBC. En este trabajo se

argumenta por qué ello es un error atendiendo al hecho de que

el prefijo dB se utiliza en los símbolos de muchas unidades

sustancialmente diferentes y a la existencia de ejemplos de

situaciones donde se sigue un criterio similar al propuesto.

Palabras clave: ISO; unidades; decibel A.

I. INTRODUCTION

The International Organization for Standardization (ISO)

recommends the symbol dB for all instances where decibel

units are used, particularly in the cases of quantities such as

the A-weighted and C-weighted sound pressure levels,

ruling out the symbols dBA y dBC.

In order to make it clear that the stated value has been A-

weighted, it is required that such condition be expressed as a

subscript in the quantity symbol, not in the unit symbol. For

instance,

Aeq

= 85 dB. (1)

The definition of the decibel itself and its symbol dB are

provided in International Standard ISO 80000-3 [1], while

the recommendation preventing the addition of any suffix

indicating the weighting is established in International

Standard ISO 80000-8:2006, which states, in a note attached

to the definition of bel and decibel,

“NOTE: The addition of a postscript to indicate the fre-

quency weighting e.g. dB(A), is incorrect. This infor-

mation should be carried by the quantity symbol, e.g.

.” [2]

This sort of recommendation is also included in the re-

port of the International Bureau of Weights and Measures

(BIPM), which states that the unit symbols should not have

any attachment, such as subscripts or suffixes, which carry

information about the quantity being represented using that

unit. Such information should be included in the quantity

symbol:

“(...) the unit symbol should not be used to provide

specific information about the quantity, and should never

be the sole source of information on the quantity. Units

are never qualified by further information about the

nature of the quantity; any extra information on the

nature of the quantity should be attached to the quantity

symbol and not to the unit symbol.” [3]

A similar criterion is expressed in the document of the

International Organization of Legal Metrology OIML D 2,

Section 1.5:

“It is not permitted to add any kind of adjective or sign to

the legal names or legal symbols of units. (For example,

electrical power is expressed in watts, W, not in electrical

watts, We).” [4]

These prescriptions are essentially correct as they are

based on an economy principle, inasmuch they discourage

assigning different symbols to the same unit. The question

is, thus, whether or not in the present case they are being

applied correctly. We will attempt to provide an answer in

the next section.

II. INADEQUATE APPLICATION OF A CRITERION

We consider that applying these prescriptions to the case

of the decibel A and similar units is a mistake that has been

repeated for a long time and should be corrected by

adopting more suitable unit symbols such as dBA, dBC, etc.,

as it was common practice decades ago and as is still

customary in a significant proportion of the scientific and

Recibido: 09/09/25; Aceptado: 21/11/25

https://doi.org/10.37537/rev.elektron.9.2.216.2025

Technical Report

Revista elektron, Vol. 9, No. 2, pp. 37-41 (2025)

ISSN 2525-0159

technical literature, as well as in normative and regulatory

documents. In what follows, an attempt will be done to

substantiate this proposal.

In the first place we are not in the same situation as in the

example provided by the BIPM publication cited above,

which states that to indicate that a given quantity U is the

maximum of an oscillating signal it would not be correct to

write

U = 1000 Vmax. (2)

Instead, it should be notated as

max

= 1000 V. (3)

In this case there is absolutely no difference between the

units represented by the symbols Vmax and V; they are,

indeed, one and the same, so it is reasonable to avoid the

use of the symbol Vmax when we already have the symbol

V to represent such unit. There is no difference whether it is

an instantaneous value, the maximum or the effective (root

mean square) value.

In the case under discussion, the situation would be

similar as if one attempted to express the peak value of a

sound pressure level as

= 102 dBpeak. (4)

to indicate that the referred value is the maximum in a given

interval. There would be no difference between the unit

whose symbol is dBpeak and the one whose symbol is dB.

There would be no difference either between the units used

to quantify sound pressure and barometric pressure.

However, the decibel A is not the same unit as plain deci-

bel. To begin with, the sound pressure level is not just a log-

arithmic form of the ratio between two arbitrary sound

pressures, as is the case of the logarithmic expression of a

gain, an attenuation or a signal-to-noise ratio, i.e., relative

quantities, which, by the way, is the only situation in which

the dB is accepted as a unit outside the SI to be used with

the SI [3]. Rather, it is an absolute unit used to express a

physical quantity univocally, just as the metre is not just a

unit to express the ratio between two lengths (e.g., the length

to be measured and the reference length) but to represent the

value of a length or distance. This is so because there exists a

specific standard reference, unlike the case of a gain, where

we can speak of the gain of a linear system regardless of the

particular values of the input and output signals.

In order to clarify this concept, the decibel, as a unit used

in acoustics, is an example of a unit to express what is

known as the level of a field quantity,

where there is a

concrete reference, in this case 20 μPa, which has been

universally accepted

and is defined in many standards, that

links the sound pressure levels expressed using this unit to

the pascal or N/m

, unit of sound pressure in the SI. Another

similar example is the dBV, used in sound systems, which

by convention uses a concrete reference equal to 1 V. We

could hardly state that the “V” that accompanies as a suffix

the particle “dB” in the symbol dBV is an attachment that

A field quantity (also called a root-power quantity) is a quantity

such as sound pressure or electric field whose square is pro-

portional to power.

For air-borne sound. In other media it may differ; for instance,

water it is 1 µPa.

onl

y informs the nature of the quantity being measured, as

would be the case in the mistaken use of symbols Vmax or

Vef to connote that we are referring to a maximum or an

effective value.

By the same token, the acoustic decibel is not just a unit

for the logarithm of the non-dimensional ratio between two

generic values of sound pressure but, rather, between the

value to quantify and a concrete reference. The fact that for

historical reasons we keep using the symbol dB for this unit

is unfortunate, since it gives rise to several misunder-

standings, including the one we are referring to. It is

tolerated because of the long tradition which dates back to

the very origins of acoustic metrology. It could be replaced

by dBp or, as is customary in many authors, dBSPL.

Now, using the same symbol dB for the A-weighted

sound pressure level is doubly confusing, as we are refer-

ring to a completely different kind of measurement, where

there is a filtering process prior to the actual measurement

which causes that in many cases where the sound pressure

levels of two signals are identical, the A-weighted values

differ, and the other way around. For instance, two sounds

of 100 Hz and 1000 Hz, both of 60 dB, have A-weighted

values of 41 dBA y 60 dBA respectively (see figure 1):

= 60 dB + A(10

0 Hz) = 60 dB − 19 dB = 41 dBA (5)

= 60 dB + A(1000 Hz) = 60 dB + 0 dB = 60 dBA (6)

whereas a sound of 100 Hz and 60 dB has an A-weighted

value of 41 dBA, the same as one of 630 Hz and 43 dB:

= 60 dB + A(100 Hz) = 60 dB − 19 dB = 41 dBA (7)

= 43 dB + A(630 Hz) = 43 dB − 2 dB = 41 dBA (8)

It is worth to note that the “A” attached as a suffix to the

unit symbol dB is not providing information on the quantity

but on its dimension, much the same as when in the case of

the symbol °C a “C” is attached to the symbol ° without the

“C” qualifying the quantity. It is attached to denote that they

are two different units since the angle and the Celsius tem-

perature are two type of quantities which are dimensionally

different. The same as in the case of the dB, here the ° has

been traditionally used for angles and temperatures, which

is also unfortunate but honors tradition. Interestingly, from

the inception of the SI the character ° is not used any longer

in the symbol of the absolute or thermodynamic temperature,

the kelvin (formerly known as degree Kelvin), whose

symbol is just K [3].

Fig. 1. Frequency response of an A-weighting filter [5].

, dB

−

f, Hz

000

Revista elektron, Vol. 9, No. 2, pp. 37-41 (2025)

ISSN 2525-0159

http://elektron.fi.uba.ar

III. THE SAME NAME FOR MANY UNITS

What is essentially happening here is that several differ-

ent units have been emerging including the prefix dB just

because there is a logarithmic process involved in the

computation of the quantity values from quantities that are

directly measured,

not because they are all quantity-

informative variants of a single unit. Using the same symbol

for units so diverse as the sound pressure level dB, the

sound intensity level dB, the spectral level dB, the gain,

attenuation or signal-to-noise dB, the voltage level dB (in its

variants dBV, dBu, dBmV, dBμV, dBkV), the electric

power level dB (dBm, dBW) is misleading. The use of a

single unit for all of them because of a fundamentalist

interpretation of the cited BIPM text is cause of confusion,

especially in disciplines such as sound technology where

several of such quantities coexist. In technology disciplines

it is frequent to find decibel quantities expressed as

−

26 dB re 1 V, (9)

where the “re” means “referred to”. This is cumbersome

since it takes a unit, a clarifying particle and a reference

quantity just to make clear what we are talking about. It

should be replaced by

−

26 dBV. (10)

IV. A SIMILAR CASE IN OPTICS

In order to see the question from another perspective we

can consider an extreme case in a similar discipline, optics,

and one of its technological branches, lighting technology.

Here we come across the case of the candela (cd), the SI unit

for luminous intensity, I

, used to quantify the brightness of a

light source ([3], p. 135). This quantity is defined as

dΦ

, (11)

where Ω is the solid angle and Φ

, the radiant flux, which is

computed in the spectral domain as

( )

λλ

=Φ



780

380

, (12)

Here d

/dλ is the radiant flux per unit wavelength and V(λ)

is the luminous efficiency, which accounts for the photopic

(daylight) spectral sensitivity of the eye (figure 2). K

is the

maximum luminous efficacy, equal to 683 lm/W. [6]

Except for the constant K

, tailored so that the luminous

intensity can be expressed in candelas, the candela would be

dimensionally equivalent to a watt per steradian (W/sr), i.e.,

the unit for radiant intensity. But the candela is, indeed, a

completely different type of unit, since the luminous inten-

sity is the result of the spectral weighting process implied in

(12). In a similar fashion as the A weighting, the luminous

efficiency weights the radiant signal according to the aver-

age visual sensitivity for each frequency (or wave length).

To some extent, both units are based on organoleptic char-

acteristics of the respective stimuli.

A notable exception is the pH (potential of hydrogen), a loga-

rithmic quantity used in chemistry to express the concentration

of hydrogen ions in a solution, which does not have a unit. It is

usually considered a scale.

Fig. 2. Luminous efficiency for photopic (daylight) vision used to compute

luminous intensity (reference color hues are only approximate)

The most noteworthy aspect is that the candela not only

has the status of an independent unit with its own symbol

included, cd, (even if it is dimensionally equivalent to the

W/sr), but it has been adopted as one of the seven fun-

damental units of the SI!

The adoption of the candela as a fundamental unit

of the

SI means an implicit acknowledgment of the fact that the

frequency weighting (filtering previous to actual measure-

ment) requires a substantially different unit with its own

unit name and symbol. The same happens regarding A-

weighted sound pressure level, and even if we do not claim

that the corresponding unit should be adopted as a funda-

mental unit, we do assert that it deserves its own symbol,

the dBA.

The discussion is bogged down by the unfortunate fact

that the symbols dB and dBA begin with the same prefix,

causing the impression that the A is just a clarifying suffix

when it is clearly not.

V. THE CASE OF THE PHON

Getting back to acoustics, let us consider the case of the

psychoacoustic quantity known as loudness level.

It is quite

a particular case in that there is not only a spectral weight-

ing but such weighting further depends on the level of the

signal [8]. A special unit, the phon, whose symbol is equal

to the name of the unit, has been introduced to express loud-

ness level. Dimensionally it would be the same as the deci-

bel, and in fact the loudness level and the sound pressure

level have the same values at 1 kHz. Since the name of the

unit does not include the prefix “dB” there has been no

objection to its use. Had it been called differently and a dB

symbol (such as dBL or dBph) been introduced, it would

have probably suffered the same fate as the dBA.

VI. THE dBA IN THE LITERATURE AND OTHER DOCUMENTS

On the other hand, the use of the dBA is customary in

many bibliographic sources, including several books, for

The fundamentalness of the candela has been challenged and

the lumen been proposed as more fundamental [7].

It is actually a psychophysical quantity which is intended as an

intermediate step for computing loudness.

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

400

450

500

550

600

650

700

750

)

Revista elektron, Vol. 9, No. 2, pp. 37-41 (2025)

ISSN 2525-0159

http://elektron.fi.uba.ar

instance [9], [10], some of which are highly regarded in the

acoustical education community, such as [11]. Other authors

prefer to notate it dB(A) ([12], [13]), which, in our opinion,

is less economical and introduces unnecessary non-alpha-

betic symbols with a mathematical interpretation that might

cause confusion (a functional notation). But all the same,

they acknowledge the need of a different unit symbol. In [14]

all three conventions (dB, dB(A) and dBA), are used de-

pending on the author of each chapter. The scientific liter-

ature contains plenty of examples using these symbol units

([15]) and the same holds for PhD theses ([16], [17]).

In documents from the World Health Organization

(WHO) dealing with community or occupational noise the

use of these units is also frequent. For instance, [18] uses

primarily dBA; [19] uses both dBA and dB(A), while [13]

and [20] use dB(A).

There are also many examples in regulatory documents,

such as in the Law 1540 from the Autonomous City of

Buenos Aires, which uses both dBA and dB(A) [21], and

the basic document of protection against noise in force in

Spain, which uses dBA [22]. The Noise Code from New

York City 0 uses dB(A), while Article 8.04 from the Code

of Ordinances of the City of Paris uses dBA [24]. The Noise

Abatement Ordinance in force in Switzerland [25] uses

dB(A).

It is noteworthy that even in relatively recent ISO stan-

dards it is possible to find cases where the dBA symbol is

used (however, they could be residual cases or cases where

the main subject of the standard is not acoustics so

reviewers might have overlooked the non-conformity with

the style requirements). Two examples are the International

Standards ISO/IWA 24:2016 [26] and ISO 16976:2023 [27].

Several ITU standards use the symbol dBA as well, such as

ITU-R BS.1771-1 0, ITU-T H.872 [29]. There exists also an

ITU recommendation, ITU V.574-5 which recommends, in

its clause 8 about special notations [30]:

“For absolute acoustic pressure level (see § 6.8)

dBA, dBB or dBC: weighted acoustic pressure level with

respect to 20 μPa, mentioning the weighting curve used

(curves A, B or C, see International Standard IEC

61672).”

The Standard IEC 60268-16, on the other side, uses an

atypical variant: dB A, with a blank between “dB” and “A”

[31]. Though we do not recommend this practice, it is

further evidence that plain dB does not convey the full story.

It is worth mentioning that until the last version of the

argentinean standard IRAM 4062 on annoying community

noise, published in 2021 ([32]), the symbol units used to

express A-weighted and C-weighted sound pressure levels

were dBA and dBC respectively. In the new version

published in 2025 ([33]) the suffixes have been dropped,

probably due to pressure to adhere completely to ISO’s

styling directives.

This is particularly problematic since this

specific standard is used as a de facto reference in many

local ordinances and regulatory documents dealing with

noise pollution, so it transcends widely the sphere of

technicians and noise control specialists. The impact is

uncertain, but there is the risk that several misunderstand-

The IRAM is the national standard organization from Argentina

and

a member of ISO.

ings could impair the effects of the standard. Indeed, some

non

-specialists could be led to believe that there has been a

change in the indicator used to assess noise pollution.

VII. CONCLUSION

We can conclude that the use, not only in academic but

also in regulatory and normative texts, of a specific symbol

(such as dBA or dB(A), with preference of the former) to

express the quantities that result from the application of a

frequency weighting is quite widespread. This reveals that

there are many specialists that acknowledge the need to

distinguish both decibel types and that the corresponding

symbology is a logical answer to that need. The cases of the

candela and the phon should be carefully considered since

they are examples of application of different criteria for a